par Autier, Philippe;Boyle, Peter;Buyse, Marc;Bleiberg, Harry
Référence Recent Results in Cancer Research, 163, page (254-63; discussion 264-6)
Publication Publié, 2003
Article révisé par les pairs
Résumé : In the three trials that tested screening with biennial fecal occult blood test (FOBT), follow-up of control patients for colorectal cancer (CRC) differed: in the Minnesota (United States) trial, the follow-up was equivalent to patients in the intervention groups, while in the Nottingham (United Kingdom) and Funen (Denmark) trials, control patients just received usual care. In the two latter trials, mortality from colorectal cancer was lower in subjects with interval colorectal cancer than in control subjects, while in the Minnesota trial, survival was equivalent in patients with interval CRC and in control patients. We examined whether better disease awareness of subjects allocated to the intervention group contributed to changes in colorectal cancer mortality observed in the FOBT trials. In the Nottingham and Funen trials, we evaluated the amount of colorectal cancer mortality reduction attributable to better survival of subjects in whom an interval colorectal cancer developed. In the Minnesota trial, we examined whether earlier detection of colorectal cancer in control subjects could explain the small (6%) reduction in colorectal cancer mortality observed with biennial FOBT. In the Nottingham and Funen trials, about one-quarter of the reduction in colorectal cancer mortality could be attributed to better awareness of patients with interval colorectal cancer. After correction for the effects of disease awareness, the absolute reduction in colorectal cancer mortality due to FOBT itself was 12% instead of 16%, and was no longer statistically significant (P>0.05). The results from biennial FOBT in the Minnesota trial published in 1993 would probably have been similar to those obtained in the Nottingham and Funen trials if disease awareness had not influenced the stage at diagnosis of colorectal cancers found in the control group. Better awareness of colorectal cancer contributes to the reduction of colorectal cancer mortality and should be encouraged. Because of a study design effect, the decrease in colorectal cancer mortality attributable to the FOBT itself is about 25% lower than that reported in the Nottingham and Funen trials. Therefore, recommending general population screening with biennial FOBT is still an open question.